Detecting the Inaudible Chatter of an Endangered Giant
How Bioacoustics Is Helping Scientists Listen to—and Protect—African Forest Elephants
In a dense tropical forest in Central Africa, a mama elephant takes a walk with her two cute calves. The calves stay close to their mother, as she is a hefty giant who shows them how to walk through and feed on the ample vegetation surrounding them. At the same time, these family members are engaged in quiet conversations using low-frequency rumbles that are mostly inaudible to humans.
They are African forest elephants, and they play a key role as architects of the forests in which they roam. However, in recent decades, their populations have declined sharply. It is therefore vital to monitor them for both scientific study and conservation—and what they “say” to each other is becoming increasingly important information.
Population Decline
A major analysis of survey data, published in the journal PLOS One, found that the population of African forest elephants dropped by as much as 62% between 2002 and 2011, and their geographic range shrank by around 30%.
The population decline can be attributed to various factors, particularly habitat loss and fragmentation through the expansion of human infrastructure and land use, and poaching for ivory and meat.
The International Union for the Conservation of Nature (IUCN), a membership group composed of government agencies and civil society organizations from around the world, publishes a Red List of Threatened Species, based on assessments carried out by ecology experts. In an assessment published in 2026, the forest elephant, Loxodonta cyclotis, is listed as “critically endangered.”
Despite their giant size, African forest elephants are among the most difficult animals to track and study. As their native habitats are the dense rainforests of Central Africa, the thick vegetation makes it almost impossible to monitor them visually. Moreover, their whispering conversations are hard to hear.
Thankfully, scientists in the field of bioacoustics have developed special technology that can detect, record, and transmit the elephants’ low-frequency sounds, making it possible to monitor the animals for scientific studies and conservation.
Before going into that, let us get to know the species.
Meet the African Forest Elephants
All elephants have features that make them easily distinguishable from other animals—particularly, their enormous size, long flexible trunk, and large flappy ears. They are mammals with high intelligence and like to socialize with one another. Biologists have given them a family name, Elephantidae.
For many years, it was thought that living elephants consist of only two species: African and Asian. They were named Loxodonta africana and Elephas maximus, respectively.
When it was recognized that African forest elephants differ from African savanna elephants (also called African bush elephants), they were initially regarded as different subspecies. Eventually, as significant differences were observed, including genetic differences, the IUCN recognized them in 2021 as two distinct species: Loxodonta africana (savanna) and Loxodonta cyclotis (forest).
Several features distinguish the African forest elephant from its savanna relative. The savanna elephant roams across the plains of sub-Saharan Africa and boasts of being the largest living terrestrial animal. The forest elephant, by comparison, is smaller in size, but it is uniquely adapted to inhabit the tropical rainforests of Central and West Africa.
It is about 8–10 feet tall and weighs around 2–5 tons, whereas the savanna elephant is 10–13 feet tall and weighs up to 7.5 tons. The forest elephant’s ears are rounder and smaller, and its tusks are relatively straight, pointing downward, and sport a pinkish tinge.
African forest elephants are herbivores, feeding mainly on fruit, grasses, leaves, and bark. They particularly enjoy tree fruits and disperse the seeds over long distances through their feces. Studies have shown that the loss of forest elephants from a part of a forest leads to the loss of trees that bear large-seeded fruit.
In addition to living in rainforests, African forest elephants also congregate and socialize at mineral-rich clearings called bai, where they can get water and the mineral supplements they need.
Elephant Talk
Elephants can produce and detect a surprising variety of sounds for communication. The sounds range from loud, broadband calls—such as trumpets, snorts, and roars—to deep, low-frequency rumbles. The rumbles include “infrasonic” vibrations, which means their frequencies extend below the approximately 20-hertz lower threshold of human hearing.
Forest elephants use their sounds to socialize, express care for one another, coordinate their movements, and find partners for reproduction.
Forest elephants use their sounds to socialize, express care for one another, coordinate their movements, and find partners for reproduction. The loud calls are often signals of threats and other emotionally charged situations. The low-frequency vibrations travel through the ground across long distances and are detected by sensitive pads in the elephants’ feet.
Katy Payne, a bioacoustics researcher, says that in 1984, she was near an elephant enclosure in a zoo, when she felt some vibrations coming from the animal. Although she could not hear it vocalizing any sounds, she decided to investigate the vibrations. Based on that experience, she cofounded the Elephant Listening Project at Cornell University.
The project has involved monitoring African forest elephants through a grid of battery-powered audio recorders mounted on trees in their habitats. The instruments can collect data over long periods without human intervention. This approach is both cost-effective and noninvasive, allowing scientists to “listen” to the elephants as they “talk” to one another. The method is called passive acoustic monitoring (PAM).
However, this approach has its challenges. In particular, the project’s archive contains hundreds of thousands of hours of rainforest sounds. The recordings include not only elephant calls but also the sounds of rainfall, birds, insects, primates, chainsaws, vehicles, gunshots, and numerous other components of the rainforest.
To analyze such enormous amounts of complex data, scientists are relying on artificial intelligence and machine-learning techniques. They first train algorithms using recordings in which elephant calls have been previously identified. They then use machine-learning systems to search new recordings to detect and classify likely elephant calls.
In one significant study, researchers assembled thousands of hours of recordings and employed machine-learning methods for elephant call detection and classification. They also developed techniques to compress the data, because conventional methods are unsuitable for the low-frequency rumbles of elephants. In this manner, scientists can now shift from sampling individual observations to analyzing entire soundscapes.
Forest elephants do not simply produce singular calls. Recent research has shown that these animals often combine their low-frequency rumbles with loud, broadband calls in their “speech.” The researchers compared it with the human use of syntax in forming sentences, noting that call combinations would lower the ambiguity of a message.
The calls of African elephants have been recorded in the context of many types of behavior. These records can be found in a large online library, The Elephant Ethogram.
Population Surveys and Conservation
The PAM approach provides a way to measure the frequency and distribution of elephant calls. It can thereby help estimate the size of forest elephant populations. Traditional surveys, such as dung counts, are difficult in dense forests and require extensive fieldwork. With PAM, the population size can be estimated without repeatedly sending large teams into remote forests.
For the conservation of forest elephants, perhaps the most immediate application of acoustic monitoring is to detect potential threats, such as poaching.
For the conservation of forest elephants, perhaps the most immediate application of acoustic monitoring is to detect potential threats, such as poaching. The microphones that detect elephant sounds can also detect human activity, including gunshots, vehicles, chainsaws, and other machinery. Such information can help park managers identify the place and time of illegal hunting and evaluate how patrols could be used effectively.
Protecting forest elephants extends well beyond protecting the species itself. They are ecosystem engineers, creating trails, dispersing seeds, and modifying the vegetation. They also help increase the aboveground biomass by promoting the growth of larger trees that lead to a higher density of carbon-capturing wood per acre. Thus, these animals contribute simultaneously to forest structure, biodiversity, and carbon storage.
For effective conservation, scientists with Cornell’s Elephant Listening Project have been collaborating with organizations such as the Wildlife Conservation Society and the World Wildlife Fund, while also training researchers and conservationists who live in Central Africa. Through such partnerships and the support of government agencies, the listening technology is more sustainable.
Where to Observe African Forest Elephants
As of 2026, the main countries these elephants inhabit are Gabon (about 66% of the total) and the Republic of the Congo (about 19%). They are also present in smaller numbers in Cameroon, the Central African Republic, the Democratic Republic of the Congo, and Equatorial Guinea, along with some in more-western Africa.
To observe African forest elephants in the wild, recommended sites to visit include the Loango National Park (Gabon), the Dzanga-Sangha Reserve (Central African Republic), and the Nouabalé-Ndoki National Park (Republic of the Congo).
*Dinshaw Dadachanji holds a PhD in Biochemistry and is a freelance writer who lives in a Maryland suburb of Washington, DC. He serves as research director of the HJ International Foundation for Environmental Peace, the institution that publishes The Earth & I.
Comments